High-Resolution Mapping of the Novel Early Leaf Senescence Gene Els2 in Common Wheat

Early leaf senescence negatively impacts the grain yield in wheat (Triticum aestivum L.). Induced mutants provide an important resource for mapping and cloning of genes for early leaf senescence. In our previous study, Els2, a single incomplete dominance gene, that caused early leaf senescence phenotype in the wheat mutant LF2099, had been mapped on the long arm of chromosome 2B. The objective of this study was to develop molecular markers tightly linked to the Els2 gene and construct a high-resolution map surrounding the Els2 gene. Three tightly linked single-nucleotide polymorphism (SNP) markers were obtained from the Illumina Wheat 90K iSelect SNP genotyping array and converted to Kompetitive allele-specific polymerase chain reaction (KASP) markers. To saturate the Els2 region, the Axiom® Wheat 660K SNP array was used to screen bulked extreme phenotype DNA pools, and 9 KASP markers were developed. For fine mapping of the Els2 gene, these KASP markers and previously identified polymorphic markers were analyzed in a large F2 population of the LF2099 × Chinese Spring cross. The Els2 gene was located in a 0.24-cM genetic region flanked by the KASP markers AX-111643885 and AX-111128667, which corresponded to a physical interval of 1.61 Mb in the Chinese Spring chromosome 2BL containing 27 predicted genes with high confidence. The study laid a foundation for a map-based clone of the Els2 gene controlling the mutation phenotype and revealing the molecular regulatory mechanism of wheat leaf senescence.


Introduction
Wheat is one of the most widely grown crops in the world, providing more than 20% of the calories and protein consumed by the human population [1]. Maintaining a high and stable yield of wheat cultivars is a huge challenge for both wheat breeders and researchers. Leaves are the main photosynthetic organ of wheat plants. A previous study showed that the photosynthetically active flag leaf can contribute 30%-50% of the assimilates for grain filling and yield [2]. The early senescence of functional leaves significantly reduces photosynthetic time and efficiency, seriously affecting grain yield and quality in wheat [3]. Therefore, discovering the genes responsible for early leaf senescence is important not only for the elucidation of the early leaf senescence mechanism, but also in terms of its applications in the development of wheat germplasms and cultivars.
Plant leaf senescence involves a complex and highly regulated process with the coordinated actions of multiple pathways [4,5]. In recent years, researchers have elucidated the molecular mechanisms of leaf senescence in model plants, and isolated and characterized several senescence-associated genes that

Development of KASP Markers Based on the 660K SNP Array
To saturate marker density in the Els2 region, we combined bulked segregant analysis (BSA) with 660K SNP arrays. A total of 10,097 SNP loci from the 660K SNP array were polymorphic between the two bulks and were located on 21 chromosomes, of which 3449 were located on chromosome 2B and the others were distributed on the other chromosomes ( Figure 3A). These results were consistent with the previous 90K SNP genotyping result [49]. Of the 3449 SNP loci located on 2B, more than 1100 were positioned in the interval 750-775 Mb according to the wheat genome assembly ( Figure 3B). Furthermore, two Els2-flanking KASP markers were located in the region 750-775 Mb (Figure 2A,C). To develop molecular markers more tightly linked to the Els2 gene, 25 SNPs located within the interval between BS00084417_51 and RFL_Contig4423_529 (763-772 Mb) were selected for conversion to KASP markers. The results indicated that nine KASP markers were tightly linked to the Els2 gene ( Figure 1B,D, Table 1).

Construction of a High-Resolution Genetic Linkage Map
To construct a high-resolution genetic linkage map of Els2, we developed a large F2 population of 838 plants. The F 2 population showed three classes of trait distribution, of which 198 were wide type plants, 419 were intermediate plants and 213 were mutant plants. The χ 2 test shows a separation ratio of 1:2:1 (χ 2 = 0.53 < χ 2 0.05,2 = 5.99), which were consistent with our previous observations [49]. The Els2-flanking KASP markers BS00084417_51 and RFL_Contig4423_529 were used to identify recombinants in the F 2 mapping population of the LF2099 × Chinese Spring cross. Only nine recombinant individuals were identified from 838 F 2 individuals. Other molecular markers, including the previously reported IP markers 2BIP06, 2BIP12 and 2BIP17 [49], were then analyzed with these recombinants only. The F 2:3 families produced by selfing the recombinant F 2 plants were used to further estimate phenotypes. In combination with the marker genotypes and recombinant phenotype, a high-resolution genetic map around the Els2 locus was eventually constructed ( Figure 2B). Among the markers flanking the Els2 gene, the KASP markers AX-111643885 and AX-111128667 were the most closely linked, at distances of 0.12 and 0.12 cM, respectively. Two KASP markers, AX-109407129 and AX-95682571, co-segregated with the Els2 gene.

Construction of a High-Resolution Genetic Linkage Map
To construct a high-resolution genetic linkage map of Els2, we developed a large F2 population of 838 plants. The F2 population showed three classes of trait distribution, of which 198 were wide type plants, 419 were intermediate plants and 213 were mutant plants. The χ 2 test shows a separation ratio of 1:2:1 (χ 2 = 0.53 < χ 2 0.05,2 = 5.99), which were consistent with our previous observations [49].
The Els2-flanking KASP markers BS00084417_51 and RFL_Contig4423_529 were used to identify recombinants in the F2 mapping population of the LF2099 × Chinese Spring cross. Only nine recombinant individuals were identified from 838 F2 individuals. Other molecular markers, including the previously reported IP markers 2BIP06, 2BIP12 and 2BIP17 [49], were then analyzed with these recombinants only. The F2:3 families produced by selfing the recombinant F2 plants were used to further estimate phenotypes. In combination with the marker genotypes and recombinant phenotype, a high-resolution genetic map around the Els2 locus was eventually constructed ( Figure 2B). Among the markers flanking the Els2 gene, the KASP markers AX-111643885 and AX-111128667 were the most closely linked, at distances of 0.12 and 0.12 cM, respectively. Two KASP markers, AX-109407129 and AX-95682571, co-segregated with the Els2 gene.

Physical Mapping and Gene Annotation of the Els2 Goal Interval
To determine the physical locations of the Els2-linked markers, the sequences of all markers anchored in the high-resolution genetic map were used to blast against the Chinese Spring genomic

Physical Mapping and Gene Annotation of the Els2 Goal Interval
To determine the physical locations of the Els2-linked markers, the sequences of all markers anchored in the high-resolution genetic map were used to blast against the Chinese Spring genomic sequence, and the relative physical positions of those markers were generally consistent with the genetic linkage map ( Figure 2B,C). The genetic region of the markers AX-111643885 and AX-111128667 corresponded to a 1.61 Mb (764913189-766527697) genomic interval in the Chinese Spring reference genome ( Figure 2C) and contained 27 annotated protein-coding genes ( Table 2). Among them, one 50S ribosomal protein L14, five cytochrome P450 family proteins, three cysteine proteinases, two serine/threonine-protein kinase, and one RECEPTOR-like protein kinase were identified.

Discussion
Mapping functional genes using ethyl methanesulfonate (EMS) mutants combined with map-based cloning has been applied in many crops; however, the high-resolution mapping of a mutant gene in wheat may become problematic and tedious without high-density molecular markers. Several approaches have been used to develop high-density molecular markers around the target gene. Comparative genomics analyses are traditional and effective approaches for marker development in a wheat positional cloning project [50]. By applying comparative genomics analyses, we previously developed five new markers linked to the Els2 gene [49]. Although release of the Chinese Spring whole genomic assembly sequences [51] would reduce the importance of comparative genomics analyses in marker development, it is still an effective method for other species without a reference genome sequence.
The SNP genotyping assays, namely 90K [38] and 660K [42], have opened a new way to undertake a more efficient and faster marker saturation of target genes in wheat [47]. Compared to the 90K SNP chip, the 660K chip has several advantages, such as higher marker density and higher resolution, as well as a better distribution on the chromosomes [42,44]. In the present study, we used the 660K chip to screen candidate markers linked to Els2 and a significant number of polymorphic SNPs were identified ( Figure 3A,B). To improve the efficiency of genotyping individuals, polymorphic SNP markers on 2BL were converted to KASP markers to genotype the F 2 population. Using this strategy, the Els2 locus was further narrowed down to a 1.61 Mb region in the Chinese Spring chromosome 2BL containing 27 predicted genes with high confidence (Figure 2B,C, Table 2). It was a clearly effective approach to exploit constantly the updated genome information and high-throughput SNP platforms for high-resolution mapping of a mutant gene in wheat. In addition to the methods used in our study, the combination of the BSA strategy with next-generation sequencing technology (RNA-Seq or exome sequencing) has been used as a mapping strategy that offers the promise of a rapid discovery of genetic markers linked to target genes in wheat [3,52,53].
The Els2 genomic interval was narrowed down to the 1.61 Mb interval of the Chinese Spring 2BL containing 27 predicted protein-coding genes. Some of the annotated genes were associated with plant early leaf senescence, including the 50S ribosomal protein L14, cytochrome P450 family proteins, cysteine proteinase, serine/threonine-protein kinase, and RECEPTOR-like protein kinase (Table 2). In rice, the gene encoding the 50S ribosomal protein L21 in the chloroplastic precursor 50S ribosomal protein may be responsible for the early senescence mutant es-t [54]. Cytochrome P450 CYP89A9, which is in charge of non-fluorescent dioxobilin-type chlorophyll catabolite accumulation, is drawn into the formation of major chlorophyll catabolites during leaf senescence [55]. Cysteine proteases are the most abundant enzymes responsible for the proteolytic activity during leaf senescence [56]. A number of cDNA clones encoding cysteine proteinases in different plants have been previously reported to be up-regulated during leaf senescence [57]. Protein kinases and autophagy-related genes were also reported to play an important role in leaf senescence [58].
In the present study, we reported our work on the high-resolution genetic mapping of the novel early leaf senescence gene Els2. First, the development of 12 KASP markers from the high-density SNP arrays enabled us to construct high-resolution genetic maps around the target locus. Second, we physically delimited the Els2 gene to a 1.61-Mb genomic region including 27 putative genes. We believe our study has made a solid foundation for the future isolation of the Els2 locus.

Plant Materials and Population Construction
The novel early leaf senescence mutant LF2099 was identified from the chemical mutagen EMS-induced library derived from the common wheat accession H261. To develop a saturated genetic linkage map for Els2, the F 2 and F 2:3 mapping populations (consisting of 838 lines) produced by crossing LF2099 to the normal leaf senescence cultivar Chinese Spring were used throughout the study. All plants were grown on the experimental farm of Northwest A&F University, Yangling, China. Leaf senescence phenotypes were evaluated by visual observation. To ensure the accuracy of the results, the phenotypes of the recombinant F 2 individuals were further validated by using the F 3 lines. The F 3 lines containing 30 seeds from every F 2 plant were planted in rows to verify the phenotypes of the F 2 plants in fields.

Combined Bulked Segregant Analysis Using the 660K SNP Array
The total genomic DNA from the parents and F 2 plants were isolated from leaves by using a cetyltrimethylammonium bromide (CTAB) method [59]. The DNA quantity was measured spectrophotometrically, and the DNA integrity was confirmed on agarose gel. To obtain closer SNP markers and saturate the targeted gene, BSA was performed to identify polymorphic markers between the early and normal DNA bulks. The DNA of the 12 homozygous plants that exhibited early leaf senescence and normal leaf senescence, based on a validation of the phenotypes of the F 3 family, was pooled to construct an early bulk sample and a normal bulk sample, separately. The F 2 bulks were genotyped with the 660K SNP arrays from CapitalBio Corporation (Beijing, China) using the Affymetrix GeneTitan System based on the Axiom 2.0 Assay Manual Workflow protocol. Allele calling was performed using the Affymetrix proprietary software according to Axiom Best Practices Genotyping Workflow. The sequences of all SNPs were used to blast version1.0 of the assembled Chinese Spring survey sequence (International Wheat Genome Consortium (IWGSC) RefSeq v1.0) to determine their physical positions [51].

Conversion of SNP Markers to KASP Markers
The polymorphic SNP markers from 90K and 660K were converted to KASP markers using the PolyMarker software (available online: http://polymarker.tgac.ac.uk) [60]. Chromosome-specific KASP markers were used to screen the parents and small population to confirm the polymorphisms before genotyping the entire mapping population. The KASP reactions were performed in a 384-well plate format following the protocol of LGC Genomics. Each KASP reaction mixtures consisted of final volumes of 5 µL containing 50-100 ng of genomic DNA, 2.5 µL of 2 × KASP Master mix (V4.0, LGC Genomics, Hoddesdon, UK), 0.056 µL of assay primer mix (12 µM of each allele-specific primer and 30 µM of common primer) and 2.444 µL of water. The assay reaction was performed in a 384 well Thermal Cycler (Bio-Rad) in the following cycling conditions: denaturation at 95 • C for 15 min, 9 cycles of 95 • C for 20 s and touchdown starting at 65 • C for 60 s (decreased by 0.8 • C per cycle), 30-40 cycles of amplification (95 • C for 20 s; 57 • C for 60 s). The amplified products were visualized by a microplate reader (FLUOstar Omega, BMG LABTECH, Germany) and analyzed by the software Klustering Caller (LGC, Middlesex, UK) [61].

Construction of High-Density Genetic Linkage Map
The linkage analysis of the polymorphic molecular markers and the Els2gene was conducted using JoinMap4.0 software [62] with default parameters and a LOD score of 3.0 as a threshold. The genetic distance was calculated and presented in Kosambi centiMorgans (cM). The genetic map was drawn with the software Mapdraw V2.1 (Huazhong Agricultural Sciences, Wuhan, China) [63].

Physical Mapping and Gene Annotation of the Els2 Region
All 14 markers used for the fine mapping of the Els2 were anchored to the wheat reference genome sequence [51]. All sequences including forward and reverse primers were searched against the Chinese Spring whole genome assembly sequences (Reference Sequence v1.0, IWGSC, http: //www.wheatgenome.org/) using the BLASTN algorithm applying default parameters. Obtained physical positions of mapped markers were visualized using the software Mapdraw V2.1 (Huazhong Agricultural Sciences, Wuhan, China) [63]. The information on the Chinese Spring genomic sequence was used to identify the genes that were included in the interval between the two Els2-flanking markers. IWGSC RefSeq v1.1 with gene annotations became available on the website https://wheaturgi.versailles.inra.fr/Seq-Repository/Annotation. Annotated genes in the target region were extracted for analyzing genes involved in plant leaf senescence.

Conflicts of Interest:
The authors declare no conflict of interest.